White matter injury (WMI) is the leading cause of life-long neurological disability in survivors of premature birth. We recently demonstrated in a large human autopsy series (Buser et al., 2012) that the major form of WMI comprises diffuse lesions and necrosis is a minor component. The mechanism of myelination failure in diffuse WMI involves arrested maturation of late oligodendrocyte progenitors (preOLs) and occurs without significant axonopathy. PreOL arrest was significantly associated with the magnitude of astrogliosis, which supports the hypothesis that factors derived from the astrogliotic lesion contribute to preOL arrest. Diffuse human WMI was enriched in astrocyte-associated hyaluronic acid (HA) that diffusely accumulated in the extracellular matrix. We found that higher molecular weight (MW) forms of HA are processed by hyaluronidases to lower MW forms that inhibit preOL maturation in vitro and delay re-myelination after experimentally-induced demyelination. We identified novel CNS expression of a GPI-anchored hyaluronidase, PH20, that localizes to preOLs in neonatal brain and which shows elevated expression within preOLs and astrocytes in white matter lesions. We propose to define novel mechanisms that involve processing of HA to bioactive lower MW forms that prevent preOL maturation in vitro and in vivo. We will test the overall hypothesis that the enhanced expression of PH20 in WMI mediates the generation of lower MW forms of HA that inhibit preOL maturation to myelinating OLs. Aim 1 addresses a series of related mechanistic questions that will define the specific size ranges of HA molecules that mediate preOL arrest in two complementary models of chronic WMI in the neonatal rat: a preterm- equivalent model of hypoxia-ischemia (H-I) and a slice culture model that displays robust astrogliosis and preOL arrest. Aim 2 integrates with aim 1 to define the regulation of tissue levels of HA in vivo in normal controls and in evolving neonatal WMI from H-I. We will quantify the developmental expression of both the HA synthases and hyaluronidases, including PH20. Aim 3 will define the effect of PH20 gain or loss of function on preOL maturation and myelination. We will quantify the magnitude of preOL arrest mediated by lentiviral-driven PH20 over-expression during normal WM development and in chronic WMI in vitro and after neonatal H-I. Complementary studies will quantify the magnitude of preOL maturation and myelination promoted by delivery of PH20 inhibitors to chronic white matter lesions arising from H-I. Aim 4 will define the temporal course of myelination failure in chronic human WMI and whether it persists or shows partial resolution. These human studies will integrate with the rodent studies in aims 1-3 to define the association of human preOL arrest with biomarkers of reactive astrogliosis (GFAP and CD44), PH20 expression, and total HA levels. We expect to gain new insights into mechanisms that trigger a failure of normal myelination in chronic WMI, and, thereby, be positioned to develop new strategies to promote normal myelination in survivors of premature birth.